The Study On The Antenna Selection Scheme Of MIMO Systems Under The Correlated Channels

Multiple-input multiple-output (MIMO) system has attracted great interesting recently since it can improve the channel capacity and reliability of wireless communications. Most researches on MIMO have made the assumption that fading channels between antenna pairs are independent. However, space correlation between antennas usually exists in real environments due to small interval between antenna elements or poor scatterings. So the space correlation is considered in this paper to meet the actual wireless communication environments. Reducing the complexity is another important research issue for MIMO systems. As we know, antenna selection scheme is the most efficient technique to reduce the system complexity and the cost of the RF as well as the effect of the serious channel correlation. Therefore the following issues about antenna selection are investigated in this thesis.Since channel capacity and symbol error rate (SER) are two basic critria to evaluate the performance of the wireless communications, it is necessary and significant to give the closed form of the channel capacity and SER. In this thesis, the closed-form channel capacity and SER are studied under the space correlation fading channels. And the theoretical results are validated by using the MATLAB simulation. The simulation results reveal that our theoretical analysis is quite precise.For the antenna selection systems, the antenna selection algorithm is one of these key problems. There are many existed selection algorithms up to now. But they are mostly only based on maximum channel capacity or minimum SER, or maximum diversity and so on. However, these algorithms can't simultaneously ensure two or more elements optimum. In this thesis, an algorithm based on maximum channel capacity and minimum symbol error rate is proposed, which is divided into two steps separately. First, part antennas are chosen to meet the maximum channel capacity criteria. Second, other antennas are chosen to meet the minimum SER criteria combined with the first chosen antennas. Then, corresponding simulations are made to validate our theoretical analysis.In most researches, it is assumed that the channel state information (CSI) is available at the receiver. However, it is ineluctable that exist estimation errors caused by Gaussian noise and the speed of mobile (Doppler frequency). In this thesis, we analyze the performance of antenna selection systems with the imperfect CSI and derive the exact bit-error-ratio (BER) expression under Rayleigh fading channel. The corresponding simulations are carried out to illustrate the impacts of Gaussian errors and Doppler spread on the performance of antenna selection systems and to prove the theoretical analysis. The best pilot length is also discussed and validated by the simulations. Finally to avoid the complicated channel estimation, a differential modulation/ demodulation scheme jointed with antenna selection at the receiver is proposed.